Populism and the scales of statehood. Localism and populist attitudes in Western Europe

Abstract

The rise of populism in Western Europe is often portrayed as a reaction to globalisation and supra-national integration processes. However, the domestic-international divide is only one aspect of the scalar organisation of government. In this article, we explore the relationship between populist attitudes and orientations towards state scales more generally. Drawing on a representative survey of 4033 citizens in Britain, France, Germany and Switzerland, we show that populist attitudes are linked to preferences for those state territories viewed as ‘closer to the people’ not only in a metaphorical but also in a scalar sense. The results suggest that the rise of populism should not only be considered a response to a crisis of party government in a context of globalisation but also as a response to a crisis of national statehood.

Introduction

Populist parties have been on the rise in Europe for more than two decades. Since the millennium, elections yielded record-highs for populist parties in many European countries, not only in those where populism has been looming since the 1990s (e.g., France, Switzerland, Austria, the Netherlands, or Belgium) but also in countries in which populism is a relatively new phenomenon (e.g., the UK, Sweden, Greece, Spain, as well as Germany). The extant literature on the topic explains the rise of populist parties as a mobilisation of the deepening cleavage between winners and losers of globalisation (see Kriesi et al., 2006; Kriesi, 2014; Hooghe and Marks, 2018). Populism, both right-wing and left-wing, is essentially viewed as a backlash to globalisation (Rodrik, 2018), portending a critique of free trade and/or de-nationalised policy-making. In this perspective, populism is seen to thrive on the tensions between the supra-national and the national spheres of policy-making, pitting advocates of international integration against advocates of national sovereignty (Basile and Mazzoleni, 2020). In Europe, European integration is particularly at stake: the question as to whether and how much nation-state sovereignty should be ceded to (or recovered from) the European Union is at the core of populist mobilisation.

While it is definitely a core feature of populist mobilisation in Europe today, the domestic-international divide is arguably only one aspect of the more general question about the scalar organisation of government. Modern states – be they federalist or unitary – encompass a multitude of institutions at different territorial layers, each with its own tasks, competencies and more or less political autonomy. The definition of their relations to the national state has been a source of ideological and political struggles ever since. The long history of state-building in Europe is replete with violent conflicts over the attribution of powers between the periphery and the centre, i.e., between local fiefdoms and the national state (Badie and Birnbaum, 1994). However, the relationships between national and sub-national layers of governments have been on the political agenda in many European countries more recently as well. On the one hand, regionalist movements have increasingly questioned the legitimacy of the nation-state in several European countries (see Fitjar, 2009). Catalunya, the Basque Country, Scotland, and Corsica are just some of the more widely known examples. On the other hand, less spectacularly but more successfully, advocacy of regional and local autonomy has sparked institutional reforms leading to devolution or decentralisation of power since the 1980s in many countries (Le Galès, 2021), thereby strengthening the regional (Hooghe et al., 2010) or the local levels of government (Ladner et al., 2016).

But what are the ideological and political positions related to the scalar organisation of government? And how do they relate to populism? The rise of populism in contemporary democracies denotes a crisis of representation – not only of representative institutions but also of political parties as the main vectors of political representation (see Mair, 2009; Caramani, 2017; Kriesi, 2018). Representation, however, is not uniscalar. It should rather be conceptualised as “compounded representation” (Brzinski et al., 1999), i.e., as the result of mobilisations that feed on political identities at various territorial scales. Political identities are like Chinese boxes: not only nations but also sub-national territories – such as regions, counties, municipalities or even neighbourhoods – provide reference points for individual political identification. But the relationship between citizens’ scalar identification and populist attitudes remains under-explored to date: dominant conceptualisations of populism, by and large, convey the de-territorialised perspective brought about by “methodological nationalism” (Wimmer and Glick Schiller, 2002). This article seeks to contribute to an emerging debate on the salient territorial dimensions of populism (see the next section for details) focusing more particularly on the local scale of government. Drawing on an analysis of survey data in four West European countries (Britain, France, Germany and Switzerland), we explore the relationships between citizens’ orientations towards different levels of government and the populist attitudes they harbour. Besides the expected relationship between populism and anti-Europeanism, we find clear evidence for a link between localism and populism. We thus argue that populism not only pits the international against the domestic sphere but in fact, portends a penchant for those governmental scales that are viewed as ‘closer to the people’ not only in a metaphorical but also in a scalar sense. It can therefore be assumed that the continued strength of populist parties and movements in many countries will further increase the salience of questions related to the scalar organisation of state power.

Populism and the scales of statehood

Given its widespread and often strategic use in the public debate, usually to denigrate political opponents, populism used to be a highly contested concept. Nevertheless, scholars increasingly agree on a common – ideational – definition of the term. Following earlier writings (see notably Canovan, 1999; Mény and Surel, 2002; Mudde, 2004; Mudde and Rovira Kaltwasser, 2013; Akkerman et al., 2014; Kriesi, 2014, 2018), populism can be defined as an ideology that considers society to be separated into two antagonistic and internally homogenous groups – the ‘corrupt elite’ and the ‘virtuous people’ – and that conceives politics as an unrestricted expression of the sovereignty of the people. This definition views populism not just as a kind of rhetoric, style or strategy, but as conveying substantive messages: anti-elitism (condemnation of the corrupt elite), people centrism (exaltation of a virtuous and homogenous people), and unrestricted popular sovereignty (Kriesi, 2018: 7). As a “thin ideology” (Mudde, 2004), the populist message can be easily combined with other ideologies. For example, right-wing populists draw on nationalism and/or nativism to defend particular cultural or ethnic communities, while left-wing populists focus on the socio-economic situation of lower classes whose interests they seek to advocate (Mudde and Rovira Kaltwasser, 2013). But how do these components of populism relate to ideological categories involved in the debate about the scalar organisation of the modern state?

The politics of scale in Western Europe

Geographical scales such as the local, the regional, the national or the global, are not something primordial but must be conceived as socially constructed and politically contested (see Smith, 1993; Delaney and Leitner, 1997; Brenner, 2002; Brenner, 2004). The organisation of state scales, as well as the articulations between them, can thus not be viewed as taken for granted or politically neutral. Instead, the scalar organisation of the state should be seen as bearing a substantial political project, thereby fixing a particular framework of social, economic and political regulation.

According to King and Le Galès (2017), state organisational structures have been refashioned twice in the 20th century. In the first phase, starting roughly after the Second World War, Europe (but also North America) experienced ‘a supple combination of state action and managed economic activity’ (King and Le Galès, 2017: 14) under a Fordist accumulation regime. State action was characterised by expanding public expenditures, redistribution, investment in infrastructure, as well as standardisation of rights and norms. These programmes led to the Keynesian welfare states of the trente glorieuses, as well as to the reconstitution of statehood at the national scale as the most relevant locus for the unleashing of the ‘unifying energy’ (King and Le Galès, 2017) of the state in this period. Many Western states also engaged in centralisation, as ‘effectively to redistribute income and resources across classes and regions necessitates centralised coordination’ (King and Le Galès, 2017: 17). The economic crises of the 1970s, together with the transformations towards post-Fordism and financialisation, induced a second, neoliberal phase of state restructuring. Under the combined pressure of increasingly globalised markets and the rise of supra-national organisations, policy-making authority has shifted away from national governments. Statehood has become more and more de-nationalised, and the distance between the state (elites) and the nation has grown.

This process of de-nationalisation entailed a shift of policy-making power away from the national government in three directions: upwards (to supra-national institutions, such as the EU), sideways (to independent regulatory agencies and private governors), as well as downwards towards sub-national authorities (Kübler, 2015). Downwards de-nationalisation has brought about profound changes in the organisation of sub-national territories, as well as altered the patterns of intergovernmental relations within nation-states across Europe in the last 40 years (Goldsmith and Page, 2010). Traditional unitary states, such as France and the UK have devolved power to newly created regional entities. Belgium embraced federalism, and decentralisation reforms have moved Spain and Italy in a federalist direction as well. And in traditional federations such as Germany, Austria or Switzerland, the federate entities have been strengthened. What is more, the EU’s approach to regions as targets and partners in its cohesion policy has contributed to a further reinforcement of Europe’s subnational governments at the expense of the national state (Goldsmith, 2002). Besides the regions, local governments have also reaped the benefits of the changing intergovernmental relations. European cities, in particular, have gained power and autonomy in the process (Le Galès, 2002), but local governments across Europe more generally have seen a substantial increase in autonomy since the 1990s (Ladner et al., 2016), even if there are significant variations across countries regarding the extent of this trend. All in all, state authority at the subnational scale has increased during the 20th century: not only in the four countries under scrutiny in this article but in Europe more generally.

In the wider perspective of the construction of state scales as political and thereby inherently conflictual processes, these changes have been viewed as a response to ‘the new socioeconomic conditions and constraints of the post-Keynesian epoch’ (Brenner, 2002: 4). In the Fordist-Keynesian era, local and regional state levels mainly operated as ‘managerial agents of nationally scaled collective consumption programmes’ (Brenner, 1999: 440) devised by highly centralised and bureaucratised states that converged around the national scale as their predominant organisational locus. This is no longer the case: in the current regime of post-Fordist and globalised capitalism, local and regional authorities serve as ‘entrepreneurial agencies’ whose major goal is to ‘enhance the locational advantages and productive capacities of their territorial jurisdictions as maximally competitive nodes in the world economy’ (Brenner, 1999: 440).

The reconfiguration of the state, as well as the reconstitution of statehood at sub-national scales also portends a shift in the substance of state policies. These are no longer geared towards integration of social groups and territories via centralised redistribution and service provision. Public policies in the reconfigured and de-nationalised state are devoid of ‘unifying energy’: they are vastly incapable of countering the rise of social inequalities. For example, the increasing manifestation of such inequalities in territorial terms has not precluded European states from gradually reducing regional development, ‘signalling further policy neglect and social decline for the abandoned backward areas’ (King and Le Galès, 2017: 27).

In sum, scale politics in European countries not only involved a strengthening of the sub-national at the expense of the national scales of government. The gradual re-constitution of statehood at the sub-national scale also epitomised the (neo-liberal) retrenchment of state policies geared towards redistribution of income and homogenisation of society. With the strengthening of subnational scales of government, the European state at the same time lost ‘its role as a unifying energizer’ (King and Le Galès, 2017: 16) of national societies.

The populist centre-periphery perspective

Scholars studying the surge of populist votes in many European countries have argued that the success of populist parties is a consequence of the inability or unwillingness of states to counter rising inequalities as a fallout of globalisation. Populist parties have indeed been viewed as mobilising the ‘losers of globalisation’ (Kriesi et al., 2008) not only in the economic but also in the cultural sense. Their electoral success in the regions threatened by decline (see Emmenegger et al., 2015), in peripheries no longer served by redistributional policies or infrastructures (see Schraff, 2019), but also in the downtrodden places of otherwise thriving urban regions (Sellers et al., 2013; Crulli, 2022) suggests that they indeed mobilise a regional or local electorate that has suffered from the weakened capacity of states to address and redress inequalities. The populist narrative thus often conveys a centre-periphery perspective that pits the virtuous (regional or local) people against corrupt (national and/or supra-national political or socio-economic) elites.

The link between populism and regionalism is a good illustration here. It is particularly obvious in cases where today’s populist protagonists have origins as ethno-regionalist parties, such as Italy’s Lega Nord or Vlaams Belang in Belgium (Rooduijn, 2018), but also Switzerland’s Lega dei Ticinesi (Mazzoleni, 2005). But populism can be linked to regionalism more generally, as Heinisch et al. (2020) forcefully show. Indeed, regional divisions across European countries allow populist actors to engage in identity politics, in the sense that some regions may “constitute an alleged ‘heartland’ containing the true representatives of the people”, whereas other regions may “be seen as collective elites whose alien values and novel lifestyles present challenges to traditional society” (Heinisch et al., 2020: 2). Heinisch et al. find several cases of regionalist parties across Europe that have developed populist claims. Examples beyond the usual suspects (i.e., the Lega Nord, the Lega dei Ticinesi, as well as Vlaams Belang) are the regionalist parties in Catalonia, the Scottish Nationalist Party or Plaid Cymru in the Celtic peripheries of the UK, but also the Alternative für Deutschland with its identitarian roots in former East Germany. In these cases, criticism of nation-state elites is a key characteristic of regionalist populism in whose discourse the enemies tend to be primarily derived from the national ‘others’. As Heinisch et al. emphasise, regionalist populism has implications for territorial politics: “grievances channeled through populist discourse are particularly effective when there are perceived disparities between the allocation of powers/resources and territorially-based identities” (Heinisch et al., 2020: 288). This assumption resonates well with the findings of a study in Spain suggesting that the electoral success of the left-wing populist party Podemos was crafted with a programmatic emphasis on the regional at the expense of the national scale (Rodriguez-Teruel et al., 2016). In a move to adapt their populist messages to the complex, multi-leveled and regionalist identities in Spain, Podemos demonstrated sympathy and support for the secessionist movements in Catalonia, Galicia and the Basque Country, and was thereby able to garner electoral support from voters favourable to decentralisation.

But the link to (ethno-nationalist) regionalism is not the only way in which populism’s ‘geographical lexicon’ (Chou et al., 2022: 132) plays out. Recent scholarship has indeed highlighted the commonalities between populism and localism – understood as an attitudinal concept that emphasises a largely positive view of the local as a scale of social, economic and political organisation (Ashton, 2010; Pied, 2011; Evans et al., 2013; Fitzgerald, 2018). As Strebel (2019: 86) points out, localism entails not only a promise of a more direct and unmediated way of citizen participation and elite control. It also conveys a managerial rationale of efficiency and service quality as localised production of public goods promises a closer match with citizens’ preferences than national production. While localism is compatible with more general ideologies about political organisation such as federalism or liberalism, it is obviously also compatible with populism. Indeed, populist narratives of the ‘ordinary people’ often entail references to local communities threatened by activities of the ‘elite’ accused of having lost touch with the base. Hence, ‘a major political expression of populism is advocacy of direct and very local democracy’ such as ‘the town hall meetings and citizen assemblies’ (Lauglo, 1995: 13). This relates back to Rousseau’s idea of common determination of the popular will in frequent assemblies, which best works under conditions of smallness and proximity. As Cochrane nicely shows, there are strong affinities between populism and localism: localism ‘seems to offer an escape from the stranglehold of traditional politics and appeals to the common sense of ‘ordinary’ people’ (Cochrane, 2016: 909), since the local tends to be considered as a ‘place where people come together more or less naturally’ (Cochrane, 2016: 910) portending the romantic view that places identity becomes more important than other forms of identity. What is more, the local has frequently been imagined as an ‘anti-bureaucratic metaphor’ (Cochrane, 2016: 910) and as a bulwark against an overreaching state, thereby also potentially appealing to populists who advocate a breaking-up of existing power relations dominated by the elite they seek to denigrate.

Localism and populism

However, the relationship between populism and localism is not limited to ideological affinities in the construction of a political narrative. In Latin America, many populist movements have long used localism for grassroots mobilisation to organise constituencies as a counterweight to entrenched power structures (Knight, 1998; Roberts, 2006). But also in the Global North, populism and localism have been strategically linked for electoral purposes. As Wills (2015) has argued, the adoption of a localist agenda by mainstream parties in the UK suggests that localism has been an electorally successful message for populists. Intending to counter the challenges by populists, leading politicians from the three UK mainstream parties – Conservative, Labour and Liberal Democrat – have ‘adopted localism as their key political agenda for post-election reform’ as a ‘product of popular, and populist, disillusionment with the nature and practice of mainstream politics’ (Wills, 2015: 188), and as an effort to engage and reach out to the people lost to populists.

The mechanisms behind these electoral effects of localism have been explored by Jennifer Fitzgerald in her recent study on radical right voting in Western Europe (Fitzgerald, 2018). Drawing on an analysis of electoral survey data, particularly in Switzerland and France, she finds that localism – operationalised as strong feelings of attachment to one’s locality – is indeed positively associated with support for radical right parties at the individual level. Her explanation for this finding rests on two arguments. On the one hand, and echoing the work on glocalisation (see Swyngedouw, 2004), she argues that economic, social and political localism is on the rise as a reaction to societal modernisation trends: in a context of globalisation and de-territorialisation, small-scale communities can offer a feeling of belonging and a sense of place. On the other hand, she argues that radical right parties are able to reap electoral benefits from these rising localist sentiments because they (a) ‘applaud a traditional version of the community, warning voters that their local areas are threatened by encroaching state authorities, supranational governance, ethnic diversity, and lack of economic protections’, (b) ‘campaign on a pro-devolution platform, promising to guard or enhance local autonomy’ as part of a claim in favour of sovereignty and territorial integrity, and (c) emphasise local attachments as part of the nativism they propagate (Fitzgerald, 2018: 11). Therefore, individuals with the strongest sense of belonging to their localities find the programmes of radical right parties particularly appealing and tend to vote for them. Very much in the same vein, a study by Arzheimer and Bernemann (2023) who investigate the geographies of populist radical right sentiment in Germany, finds that nativism, right-wing authoritarianism and populist attitudes are significantly related to localism.

However, while these studies develop a convincing ‘localist theory of radical right voting’ (Fitzgerald, 2018: 7), and present evidence for a (weak) link between localism and populist radical right sentiment (Arzheimer and Bernemann, 2023) they do not strictu sensu, offer an explanation about the link between localism and populism in general. This question, indeed, remains open to date. Chou et al. (2022) have very recently formulated a set of assumptions intended to systematise our understanding of the intersections between localism and populism. These assumptions rest on the basic argument that localism plays a role in both right-wing and left-wing populism:

• Right-wing populism is related to localism in three ways. First, there is the nativist theme, where the good people of the local communities need to be protected against the out-of-touch elite in the capital or even more foreign places. Second, the local is invoked as a traditional version of community that often comes with the romanticisation of rural life and the idealisation of honest, hard-working locals versus the conditions dictated by the supra-local elite and state authorities. Third, there is advocacy of local autonomy and devolution.

• Left-wing populism’s conceptualisation of the local emphasises two points. First, it thematises local suffrances as an outflow of global (capitalist) power, advocating that locals should participate more in collectively governing communities. Second, it promotes microcosms of the inclusive and egalitarian communities and calls for action whereby people unite to take back control over their community from the elites, represented by the wealthy upper class or global corporations.

In sum, Chou et al. propose that ‘nativism is the overriding ideology that discursively “fills” “the local” for right-wing populists, whereas for left-wing populists, a zeal for participatory politics “on the ground” seems to play the same role’ (2022: 135). They thus make a strong case for the argument that localism fosters populism. However, empirical evidence for this conjecture is as yet unavailable. The aim of the present study precisely is to fill this gap, by testing the hypothesis that localism is positively associated with populist attitudes. In doing so, we conceive localism not as an absolute, but as a relative measure, i.e., as a scalar orientation that prefers the local over the national level of government and politics. Indeed, drawing on the concept of ‘nested identities’ (see Herb and Kaplan, 1999) we argue that territorial identities are negotiated within a hierarchy of geographical scales and should therefore be measured in relation to each other.

Data, variables and method

We test this hypothesis empirically on the basis of individual-level data collected via a representative, mixed-method survey of citizens aged between 18 and 75 years in four European countries. The survey covered, among other items, both populist attitudes, as well as scalar orientations – both conceived as continuous measures ¹ (see operationalisation of variables in the next section as well as Table A1 in the online Appendix for question wordings).

Sampling and survey

The survey on Democratic governance and citizenship in Europe (DemGovCit) collected individual-level data on political values and behaviour, as well as on attitudes and assessments of various aspects of democracy and governance. The study was approved by the ethics committee of the Faculty of Philosophy of the University of Zurich on 26th of June 2015, and the survey was fielded in fall of 2015. Respondents were recruited offline on a randomised basis. Due to national differences in accessibility of address data, recruitment procedures differed slightly across the four countries. While official registries could be used for initial recruitment of respondents in Switzerland via ground mail, Computer Assisted Telephone Interviews screening with random digit dialling (including mobile phone numbers) was used to select potential respondents in Germany, France and Britain. Respondents were incentivised to participate in the survey: 10 EUR in Germany and France, 10 GBP in Britain, and 10 CHF in Switzerland. Interviews were administered in mixed mode: online questionnaires as the standard procedure, and paper questionnaires (including a prepaid return envelope) were sent to respondents without private internet access upon request. The survey was conducted by commercial providers MIS Trend (in Switzerland) and TNS Infratest (in France, Germany and Britain).

Interviews were completed by a total of 4033 respondents: France (N = 1031), Germany (N = 1111), Switzerland (N = 924), and Britain (N = 977). To correct sampling bias, sampling and post-stratification weights were calculated based on household size and number of phones (to correct for varying selection probabilities into the sample in the French, German and British samples) and on age, gender, education, employment status, and region (to correct for non-response bias). The analysis presented hereafter uses weighted data.

Dependent variable: populist attitudes

The aim of this paper is to gauge the influence of respondents’ perceptions of and attitudes towards various territorial scales of state organisation on populist attitudes. Drawing on the definition of populism elaborated by Mudde (2004), we conceive populist attitudes as a function of three dimensions: anti-elitism, people-centrism and demands for popular sovereignty. Following the scale developed by Schulz et al. (2018), we operationalise these three dimensions on the basis of (a) two questions gauging a respondent’s perception of the political elite (anti-elitism), (b) four questions regarding a respondent’s perception of the people as a virtuous and homogenous group (people-centrism), as well as (c) two questions about the respondent’s attitudes towards direct participation of citizens in political decision-making (popular sovereignty). ² For each of these three dimensions, we calculate arithmetic means of the associated items which then yields three indices for anti-elitism, people-centrism and popular sovereignty. We also calculate an overall measure of populist attitudes. For the latter, we acknowledge the important argument made by Wuttke et al. (2020) that the key characteristic of populism is its conceptualisation as an attitudinal syndrome with non-substitutable sub-dimensions and therefore follow Mohrenberg et al. (2019) in using the geometric mean of the three preceding composite indices. This is achieved by the following formula:

$${\rm{Populism\;}}\left( {{\rm{overall}}} \right) = \;^3\hskip-8pt\sqrt{{{\rm{Anti \hbox- elitism}}*{\rm{People \hbox- centrism}}*{\rm{Popular\;sovereignty}}}}$$

Using the geometric rather than the arithmetic mean makes sure that the overall index for populist attitudes equals zero when one of its three components (scaled from 0 to 4) is zero.

Independent variables: relative scalar orientations

The main aim of our study is to explore the relationship between populist attitudes and localism at the individual level – both conceived of as a matter of degree rather than principle. More precisely, we seek to test the hypothesis, that populist attitudes in the electorate are associated with localist orientations, where localism is conceptualised as ‘a set of practices, claims and discourses, that render “the local” politically salient, that is, an ideational politics of place’ (Chou et al., 2022: 131), and measured as an attitude, i.e., a feeling of belonging or attachment (Fitzgerald, 2018; Arzheimer and Bernemann, 2023). Our contention is that scalar orientations within the domestic sphere are of similar importance for populist attitudes as scalar orientations with respect to the national-international divide.

Operationally, and following earlier work on citizens’ scalar orientations (Kübler, 2018), we use measures for respondents’ attachment to different territories to calculate relative scalar orientations. More precisely, local scalar orientations are calculated by subtracting the values of respondents’ answers to survey questions about emotional attachment to their country, from the values of their answers to questions about their attachment to their place of residence. This yields a measure of local scalar orientation that runs from minus 10 to plus 10. Negative values of this variable thus denote stronger orientations towards the national scale, while positive values denote that respondents are more oriented towards the local scale. As a corollary, scalar orientations on the national-international divide are calculated by subtracting the values of respondent’s attachment to their country from the attachment to Europe. Here, positive values express a more European orientation, whereas negative values denote an orientation towards the national scale.

Control variables

In terms of control variables, we have to consider that both populist attitudes, as well as scalar orientations, are likely to differ across the four national contexts in which our study was conducted. As the aim of this article is to explore the relationships between populist attitudes and localist orientations in general, we do not formulate specific hypotheses about how these differ between the countries under scrutiny. However, we need to take into account that national specificities could play a role and therefore include country-dummies to control for these effects. At the individual level, we include socio-demographic control variables (age, gender and education), as well as attitudes towards immigrants to control for nativist leanings. Indeed, in the four countries under scrutiny, most populist parties are located on the right of the political spectrum, which makes a strong association between nativism and populism very likely (Rooduijn, 2018). To test whether scalar orientations are linked to populist attitudes across the ideological spectrum and for supporters of different parties, we interact scalar orientations with respondents’ left-right self-placement and with their party identification. For the latter, we classified respondents into four groups: no party identification, other party identification, left-wing populist party identification, and right-wing populist party identification. We classified parties as left- or right-wing populists based on the PopuList dataset (Roodujin et al., 2023). While the main populist parties in the four countries studied here are located on the right side of the political spectrum, there are also two well-established left-wing populist parties in France and Germany. ³

Results

Populist attitudes and scalar orientations: bivariate relationships

To get a sense of how scalar orientations are linked to populist attitudes, we begin the empirical analysis section with descriptive evidence on how ‘populist’ and ‘non-populist’ respondents differ in their scalar orientations. To do so, we follow Wuttke et al. (2023) and generate two groups of respondents: the ‘populists’ that score 3 or higher on all three dimensions of populism (on a scale from 0 to 4) (N = 314) and the ‘non-populists’ that score lower than 3 on at least one dimension (N = 3719). Figure 1 shows mean levels of attachment to the local, the national, and the European scale as well as the average local and European scalar orientation for populists and non-populists.

Figure 1. The scalar orientations of ‘populists’. (a) Mean attachment levels to different scales. (b) Local and European scalar orientations.

These two figures provide several interesting insights. First, Fig. 1a shows that populists are on average significantly more attached to the local and less attached to the European level than non-populists. They also tend to be more attached to the national level than non-populists but this difference is not statistically significant. Second, Fig. 1b shows that populists are on average more oriented towards the local scale than non-populists – i.e., they have less negative values on the local scalar orientation variable and more negative values on the European scalar orientation variable. ⁴ On the one hand, this suggests that ‘populists’ are indeed relatively more oriented to the local and less to the European scale than non-populists. On the other hand, it also shows that this difference is mainly one of degree and not of principle (see Schäfer, 2022): it is not the case that populists all feel more attached to the local than to the national level in absolute terms, but rather that they feel relatively more attached to the local than to the national scale than non-populists.

As expected, populist attitudes are strongly related to preferences for populist right-wing parties according to the PopuList classification (Rooduijn et al., 2023). Respondents declaring affinities with populist right-wing parties score significantly higher on all three dimensions of populist ideology, as well as on the overall index of populism (see Figure A1 in the online appendix), thereby confirming previous evidence (Rooduijn, 2018). For supporters of left-wing populist parties, the results are less clear. They do not score higher than supporters of other political parties on the overall populism measure – which mainly has to do with their low scores on the people-centrism dimension. Yet, they score significantly higher on the anti-elitism and the popular sovereignty dimension than supporters of other political parties.

Relationships between populist attitudes and scalar orientations

An exploration of the correlations between the three dimensions of populist attitudes and the overall index of populism on the one hand, as well as the two different measures for respondents’ scalar orientations on the other hand, corroborates the assumption that there is a ‘scalar dimension’ to populism (Table 1). While a European scalar orientation is negatively correlated with all three dimensions of populism as well as the overall index of populism, the reverse is true for a local scalar orientation. However, although statistically significant, not all of these correlations are equally strong. A closer inspection of the coefficients shows that a European scalar orientation is less strongly correlated to anti-elitism than to the other dimensions of populism or to the overall index.

Table 1. Correlations between populist attitudes and scalar orientations (Pearson’s correlation coefficients)

Notes: weighted data, levels of significance: *p < 0.05, **p < 0.01, ***p < 0.001.

In order to test the robustness of the association between scalar orientations and populist attitudes, multi-variate regression analysis was used. More precisely, four OLS models were estimated, regressing the two measures for scalar orientations on the four indices for populist attitudes. Besides country-dummies that control for unobserved country-level confounders, a number of individual-level controls were used to filter out confounding effects of socio-demographics such as age, gender and education, as well as nativist leanings. Regarding the main question of interest in this study, the multivariate results straightforwardly corroborate our hypothesis that scalar orientations are associated with populist attitudes (Fig. 2). Local scalar orientations (where higher values indicate a relatively stronger attachment to the local as compared to the national level) are positively related to all dimensions of populism, as well as to the overall measure of populism, whereas European scalar orientations (where higher values indicate comparatively stronger attachment to the European than to the national level) are negatively associated with populist attitudes. While the results regarding the European orientation confirm previous evidence, the findings on the effect of a local scalar orientation are a new insight: those who feel relatively more attached to the local than to the national level also hold stronger populist attitudes. This extends and broadens recent findings on the positive relationship between local attachment and support for populist radical right parties and ideas to populist attitudes in general (Fitzgerald, 2018; Arzheimer and Bernemann, 2023).

Figure 2. The relationship between scalar orientations and populist attitudes.

Note: OLS-regression coefficients with 95% confidence intervals. Full regression model in Table A2 in the Appendix.

In order to assess if these findings are affected by unobserved country effects, we applied a Jackknife procedure and re-estimated the regression model on the overall index of populist attitudes, excluding one country at a time. As the results show (see Figure A2 in the appendix), the findings reported above are indeed robust to the exclusion of particular countries from the analysis.

We further probed this result by analysing whether the association between scalar orientations and populist attitudes differs depending on political ideology. Existing studies on localism and populism have predominantly focused on populist radical right parties (e.g., Fitzgerald, 2018, Arzheimer and Bernemann, 2023). It might, thus, be the case that the association is mainly present among citizens on the (far) right of the political spectrum. To test this, we interact the scalar orientation measures with two different variables that capture citizens’ ‘thick’ political ideologies: their left-right self-placement and their party identification. This allows us to assess whether the relationship between scalar orientations and populist attitudes depends on political ideology.

To assess whether respondents’ ideological position moderates the association between their scalar orientations and their populist attitudes, we calculate interaction effects. The interaction effect between left-right self-placement and scalar orientations is displayed in Figs. 3 and 4 (full model results are in Table A3 in the appendix). The relationship between local scalar orientations and populist attitudes is not significantly moderated by respondents’ ideological positions. While Fig. 3 shows a certain tendency for a stronger relationship between local scalar orientations and the overall measure of populism for respondents on the political right, the difference to respondents on the political left is not statistically significant. Across the ideological left-right spectrum, local scalar orientations are, thus, positively associated with populist attitudes.

Figure 3. The relationship between populist attitudes and local scalar orientation across the ideological spectrum.

Note: OLS-regression coefficients; shaded area depicts 95% confidence intervals. Full regression models are in Table A3 in the Appendix.

Figure 4. The relationship between populist attitudes and European scalar orientation across the ideological spectrum.

Note: OLS-regression coefficients; shaded area depicts 95% confidence intervals. Full regression models are in Table A3 in the Appendix.

The picture is somewhat different for European scalar orientations. While for the overall measure of populist attitudes, we do not find a statistically significant moderation effect of left-right self-placement, Fig. 4 shows that left-right ideology matters for the relationship between European scalar orientations and popular sovereignty as well as anti-elitism. In both cases, the association between the two is significantly stronger for respondents on the political right. When we classify respondents into different groups based on their party identification (see Table A4 as well as Figures A3 and A4 in the appendix), we do not find statistically significant differences in the relationship between scalar orientations and populist attitudes across supporters of different parties. ⁵ Overall, this suggests that the relationship between scalar orientations and populist attitudes does not depend on citizens’ political ideology. Rather, it seems that across the political spectrum, scalar orientations are related to populist attitudes.

Discussion and conclusion

The question asked at the outset of this study was whether and how particular perspectives on the scalar organisation of the state relate to populist attitudes in Western Europe. More precisely, we hypothesised that populist attitudes are negatively associated with scalar orientations that value the European scale against the national scale, and positively associated with scalar orientations that prefer the local to the national scale of government and politics. Our findings corroborate this hypothesis in a straightforward manner. All three dimensions of populist attitudes– anti-elitism, people-centrism, claims for popular sovereignty – but also the overall measure for populist attitudes in citizens, were found to be associated with scalar orientations as expected. While our findings thereby confirm earlier evidence that populism portends a scalar dimension, they also buttress the argument that this scalar dimension goes beyond the widely discussed international-domestic divide: citizens with strong populist attitudes tend to value ‘closeness to the people’ not only in the metaphorical but also in the scalar sense. This is in line with our argument that populist citizens hold scales that are closer to them more dear than non-populist citizens.

More specifically, our findings clearly show that localism is positively associated with populism. In this sense, they support the idea that Fitzgerald’s (2018: 7) ‘localist theory of radical right voting’ can be extended and generalised into a ‘localist theory of populism’ so to speak. Our analysis nicely ties in with Chou et al.’s (2022) conjecture that localism plays a role in populist attitudes, both right-wing and left-wing.

But our analysis also contributes to a better understanding of the populist phenomenon more generally. In the extant literature, the rise of populism is interpreted as resulting from a crisis of political representation propelled by two different but complementary processes (Kriesi, 2018: 14ff). On the one hand, the declining ability of political parties in most Western democracies to mobilise voters and channel political conflict has opened up new opportunities for populist protesters rallying against the supposedly privileged political class. On the other hand, the emergence of new structural conflicts has been emphasised, such as a deepening “transnational cleavage” (Hooghe and Marks, 2018) between winners and losers of globalisation processes, leading to an integration-demarcation divide in the political space, in which populists mobilise the opponents to supra-national integration to whom established parties have not been responsive. While this perspective obviously provides a convincing explanation for the populist furor against globalisation and supra-national integration, it is less evident how it relates to our findings that populism also feeds on criticism of the national scale of government with respect to the local one. In this sense, our findings strongly contradict an assumption formulated by Basile and Mazzoleni (2020) according to which populism is just national sovereignism in new bottles. Populists, we found, cherish the local rather than the national state. Our study does not provide clear-cut answers to the question of why they do so. However, our findings suggest that the rise of populism is not only a story of a crisis of party government in a context of globalisation. It is also a story of national statehood in crisis, most likely as a consequence of restructuring processes in the second half of the 20th century– aptly described by King and Le Galès (2017).

This study obviously has its limitations. Most importantly, the unavailability of geo-coded survey data prevents us from analysing whether and how the relationship between scalar orientations and populist attitudes hinges on specificities of the local context. Arzheimer and Bernemann’s (2023) findings for Germany suggest that localist attitudes affect populist sentiments independently from place-related characteristics. Nevertheless, a closer inspection of local context effects could yield insights into the co-construction of local identities and populist attitudes by populist actors – as, for example, evidenced by Volk (2022) in the case of the right-wing and anti-islamic PEGIDA movement in Dresden (East Germany). In the same vein, further research into the role of regionalist mobilisation in the relationship between scalar orientations and populist attitudes is needed. ⁶ Focusing on countries with strong and diverse regionalist parties – such as Spain – would be particularly promising to this endeavour. While these questions must remain open for the moment, we think that our findings make a strong case for the analytic potential that lies in the conceptual and methodological de-nationalisation of the study of populism.